Path planning method for leader-follower multi-vehicle formation with integrating GoT-SAC

Yue Wang; Hongwei Duan; Wei Zhong; Lu Yang; Lei He; Fulai Chai; Xiaoyang Shi

doi:10.3969/j.issn.1674-8484.2026.01.013

Path planning method for leader-follower multi-vehicle formation with integrating GoT-SAC

Yue Wang
, Hongwei Duan
, Wei Zhong
, Lu Yang^*
, Lei He
, Fulai Chai
, Xiaoyang Shi

^*Corresponding author for this work

School of Mechanical Engineering
University of Science and Technology Beijing
Tsinghua University
Beijing Institute of Technology

Research output: Contribution to journal › Article › peer-review

Abstract

A leader-follower formation path planning method was proposed through integrating the Goal-oriented Transformer (GoT) and the Soft Actor-Critic (SAC) on the Mecanum Wheeled intelligent platform, named GoT-SAC, to enhance the stability and efficiency of formation operation in unknown environments. Experimental validation was conducted in both the Gazebo environment and on a miniature physical platform. The results show that the GoT-SAC model converges within 95–100 training episodes. The average relative pose error reduces from 18 cm to 6 cm with a path-length relative-difference being below 5% compared with the manual remote-control strategy. Therefore, the proposed method achieves stable formation and efficient obstacle avoidance without relying on prior map information.

Translated title of the contribution	融合 GoT-SAC 的领航—跟随式多车编队路径规划方法
Original language	English
Pages (from-to)	122-129
Number of pages	8
Journal	Journal of Automotive Safety and Energy
Volume	17
Issue number	1
DOIs	https://doi.org/10.3969/j.issn.1674-8484.2026.01.013
Publication status	Published - 2026
Externally published	Yes

Keywords

deep reinforcement learning
leader-follower formation
multi-vehicles formation control
path planning

Access to Document

10.3969/j.issn.1674-8484.2026.01.013

Cite this

@article{8791ba469d1b4e1d816c49c8d1f91e9f,

title = "Path planning method for leader-follower multi-vehicle formation with integrating GoT-SAC",

abstract = "A leader-follower formation path planning method was proposed through integrating the Goal-oriented Transformer (GoT) and the Soft Actor-Critic (SAC) on the Mecanum Wheeled intelligent platform, named GoT-SAC, to enhance the stability and efficiency of formation operation in unknown environments. Experimental validation was conducted in both the Gazebo environment and on a miniature physical platform. The results show that the GoT-SAC model converges within 95{\textendash}100 training episodes. The average relative pose error reduces from 18 cm to 6 cm with a path-length relative-difference being below 5\% compared with the manual remote-control strategy. Therefore, the proposed method achieves stable formation and efficient obstacle avoidance without relying on prior map information.",

keywords = "deep reinforcement learning, leader-follower formation, multi-vehicles formation control, path planning",

author = "Yue Wang and Hongwei Duan and Wei Zhong and Lu Yang and Lei He and Fulai Chai and Xiaoyang Shi",

year = "2026",

doi = "10.3969/j.issn.1674-8484.2026.01.013",

language = "English",

volume = "17",

pages = "122--129",

journal = "Journal of Automotive Safety and Energy",

issn = "1674-8484",

publisher = "Tsinghua University Press",

number = "1",

}

TY - JOUR

T1 - Path planning method for leader-follower multi-vehicle formation with integrating GoT-SAC

AU - Wang, Yue

AU - Duan, Hongwei

AU - Zhong, Wei

AU - Yang, Lu

AU - He, Lei

AU - Chai, Fulai

AU - Shi, Xiaoyang

PY - 2026

Y1 - 2026

N2 - A leader-follower formation path planning method was proposed through integrating the Goal-oriented Transformer (GoT) and the Soft Actor-Critic (SAC) on the Mecanum Wheeled intelligent platform, named GoT-SAC, to enhance the stability and efficiency of formation operation in unknown environments. Experimental validation was conducted in both the Gazebo environment and on a miniature physical platform. The results show that the GoT-SAC model converges within 95–100 training episodes. The average relative pose error reduces from 18 cm to 6 cm with a path-length relative-difference being below 5% compared with the manual remote-control strategy. Therefore, the proposed method achieves stable formation and efficient obstacle avoidance without relying on prior map information.

AB - A leader-follower formation path planning method was proposed through integrating the Goal-oriented Transformer (GoT) and the Soft Actor-Critic (SAC) on the Mecanum Wheeled intelligent platform, named GoT-SAC, to enhance the stability and efficiency of formation operation in unknown environments. Experimental validation was conducted in both the Gazebo environment and on a miniature physical platform. The results show that the GoT-SAC model converges within 95–100 training episodes. The average relative pose error reduces from 18 cm to 6 cm with a path-length relative-difference being below 5% compared with the manual remote-control strategy. Therefore, the proposed method achieves stable formation and efficient obstacle avoidance without relying on prior map information.

KW - deep reinforcement learning

KW - leader-follower formation

KW - multi-vehicles formation control

KW - path planning

UR - https://www.scopus.com/pages/publications/105036025915

U2 - 10.3969/j.issn.1674-8484.2026.01.013

DO - 10.3969/j.issn.1674-8484.2026.01.013

M3 - Article

AN - SCOPUS:105036025915

SN - 1674-8484

VL - 17

SP - 122

EP - 129

JO - Journal of Automotive Safety and Energy

JF - Journal of Automotive Safety and Energy

IS - 1

ER -

Path planning method for leader-follower multi-vehicle formation with integrating GoT-SAC

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this